Against “neurobabble”

Every written token of the English word “soup” is made up of marks which look at least vaguely like “s,” “o,” “u,” and “p.”  Of course, it doesn’t follow that the word “soup” is identical to any collection of such marks, or that its properties supervene on the material properties of such marks, or that it can be explained entirely in terms of the material properties of such marks.  Everyone who considers the matter knows this.

To borrow an example from psychologist Jerome Kagan, “as a viewer slowly approaches Claude Monet's painting of the Seine at dawn there comes a moment when the scene dissolves into tiny patches of color.”  But it doesn’t follow that its status and qualities as a painting reduce to, supervene upon, or can be explained entirely in terms of the material properties of the color patches.  Everyone who considers the matter knows this too.

Somehow, though, when neuroscientists discover some neural correlate of this or that mental event or process, a certain kind of materialist concludes that the mind’s identity with, or supervenience upon, or reducibility to, or complete explanation in terms of neural processes is all but a done deal, and that the reservations of non-materialists are just so much intellectually dishonest bad faith.  In a recent online op-ed piece for The New York Times, and in an apt phrase, philosopher of mind Tyler Burge criticizes this tendency as “neurobabble,” which produces only “the illusion of understanding.”  For it is as fallacious as any parallel argument about words or paintings would be.

Now one source of neurobabble is the standard but false materialist assumption that the only dualistic alternatives to a “naturalistic” account of the mind are either Cartesian substance dualism or property dualism, with their attendant interaction problem.  To be sure, and as I have noted many times, materialists often deeply misunderstand even these forms of dualism (or at least Cartesian dualism) and direct their objections at crude straw men.  [For some examples, see this post on Daniel Stoljar, and my four-part series of posts on Paul Churchland, here, here, here, and here.  For discussion of the shallowness of materialist arguments in general, see this post on Frank Jackson and this post on (the non-shallow) Noam Chomsky.]

Still, from an Aristotelian-Thomistic (A-T) point of view, even Cartesian dualism is a modernist error, the “evil twin” of materialism.  It exaggerates the divide between mind and matter, even as materialism exaggerates their affinity.  (For A-T, many modern positions are “evil twins” in this sense – rationalism and empiricism, libertarianism and socialism, Kantian deontology and utilitarianism, and so on – each removing a genuine insight from the classical metaphysical framework in which it makes sense and then twisting it into a grotesque caricature of itself by ignoring the opposite, balancing insight.  I’ve been meaning to write up a post on that theme, but it is addressed at least indirectly in The Last Superstition.)

The A-T approach is what David Oderberg has called “hylemorphic dualism.”   Unlike Cartesian dualism, which regards a human being as a composite of two substances, res cogitans and res extensa, hylemorphic dualism regards a human being as a single substance.  But unlike materialism, which tends to regard material substances as reducible to their component parts and which is wedded to a mechanistic conception of matter that denies the reality of formal and final causes, hylemorphic dualism is non-reductionist, and regards human beings, like all material substances, as composites of form and matter.  (The view is non-reductionist despite regarding material substances as composed of form and matter, because it does not reduce them to form and matter.  A tree, for example, is a composite of a certain kind of form and matter, but the form and matter themselves cannot be made sense of apart from the tree of which they form metaphysical parts.  The analysis is holistic.) 

“Soul” on this view is just a technical term for the form of the living body.  And the view is dualist, not because it affirms the existence of the soul (plants and non-human animals have forms, and thus “souls,” but are purely material) but rather because it takes human beings to have certain special capacities that do not involve a material organ – namely, their intellectual capacities.  There is no “interaction problem” for hylemorphic dualism, though, because the soul is not (as it is for Descartes) a distinct substance which needs somehow to get into contact with a material substance via efficient causation; it is rather only a part of a complete substance – the formal cause of the substance, of which the matter composing the body is the material cause.  The relationship between soul and body is therefore not like that of two billiard balls, one of them ghostly, which have to find a way somehow to knock into one another.  It is more like the relationship between the shape of a triangle drawn on paper and the ink which has taken on the shape – two aspects of one thing, rather than two things.  Or it is like the relationship between the meaning of a word and the letters that make up the word, or the relationship between the pictorial content of a painting and the splotches of color that make up the painting.  (Probably most of my readers will be familiar with these ideas, but for those who are not, I have spelled them out in more detail in many other places, most fully in chapter 4 of Aquinas.)

One problem with many claims made for materialist reductionism, then, is that they rest on a conception of part-to-whole relations in material substances that is (on the A-T view) false across the board, not merely where the mind-brain relationship is concerned.  It is false to say that a tree is “nothing but” a collection of roots, trunk, leaves, sap, etc., even though a tree does of course have such parts.  It is false to say that a triangle is “nothing but” the ink particles that make up its lines, that a word is “nothing but” the material marks that comprise its tokens, or that a painting is “nothing but” the color patches that the painter has put on canvas, even though these objects also have the parts in question.  And it is false to say that the mind is “nothing but” a collection of neural processes, even though neural processes do indeed underlie all of our mental activities.  (You don’t have to be an A-T theorist to see this, by the way.  See M. R. Bennett and P. M. S. Hacker, Philosophical Foundations of Neuroscience for a thorough critique of the conceptually sloppy and fallacious thinking that permeates much philosophical and “scientific” discussion about the brain.)

Now, since A-T is committed to a kind of dualism, albeit of the hylemorphic variety – and since, in particular, it holds that intellectual operations have no bodily organ – it might sound surprising that I should say that “neural processes do indeed underlie all of our mental activities.”  But that is indeed exactly what the hylemorphic dualist claims.  The reason is this.  Keep in mind first of all that A-T regards sensation and imagination – those “mental” phenomena we have in common with lower animals, and which are characterized by what contemporary philosophers call “qualia” – as corporeal or bodily in nature, and in that sense entirely material.  To be sure, A-T has a different conception of matter than materialists do.  For example, A-T does not hold that the only properties of matter are those described by the modern physicist.  But the relevant point for present purposes is that A-T does not regard sensation and imagination per se as involving any sort of immaterial organs or properties, anything that survives the death of the body, or anything that distinguishes us from the brutes.

What does distinguish us from the brutes and entails immateriality is our grasp of concepts or universal ideas.  One reason conceptual thought cannot be material is that concepts and the thoughts that feature them are abstract and universal, while material objects and processes are inherently concrete and particular; another is that concepts and the thoughts that feature them are (at least sometimes) exact, determinate, and unambiguous while material objects and processes are inherently inexact, indeterminate, and ambiguous when they are associated with conceptual content at all.   There are other reasons too.  (These are issues I have addressed many times.  For a more detailed treatment, see chapters 6 and 7 of Philosophy of Mind and, again, chapter 4 of Aquinas.  Some relevant blog posts of mine can be found here and here.  And see also James Ross’s article “Immaterial Aspects of Thought” and David Oderberg’s article “Concepts, Dualism, and the Human Intellect.”)

All the same, given that the soul of which intellect is one of the powers is of its nature oriented to the body, of which it is the form, the human intellect – unlike the intellects of angels, which are akin to Cartesian immaterial substances – requires bodily activity as a necessary condition of its ordinary operation, even if it is not a sufficient condition.  For one thing, it requires that there be sense organs to generate the sensations from which “phantasms” or mental images can be derived, from which in turn the intellect can abstract concepts.   But it also (and more to the present point) requires that there be organs capable of generating phantasms or images even after sensation has ceased; that is to say, it requires the neurological processes underlying imagination.  For even though our concept of a triangle (for example) is not and cannot be identified with any image of a triangle – such an image will always have features that the concept lacks, will strictly apply only to some triangles while the concept applies to all, might be vague in certain respects, and so forth – we are nevertheless incapable of entertaining the concept of a triangle without at the same time forming an image of some sort (a mental picture of a triangle, or of the look or sound of the word “triangle,” or whatever).  

A useful analogy would be Frege’s conception of the relationship between propositions and sentences.  A proposition cannot be identified with a sentence; for instance, the proposition that snow is white cannot be identified with the English sentence “Snow is white,” because someone who spoke German rather than English could express the very same proposition by using the sentence “Schnee ist weiss.”  But neither can it be identified with any other sentence or collection of sentences, since the proposition that snow is white was true before any language came into existence, and would remain true even if every language went out of existence.  In short, propositions are not linguistic entities.  All the same, they cannot be grasped by us except by means of linguistic entities.  The proposition that snow is white is not identical with “Snow is white” or “Schnee ist weiss,” but you cannot entertain it without entertaining either one of those sentences, or a sentence of some other language.  As Frege put it in his classic paper “The Thought”: “The thought, in itself immaterial, clothes itself in the material garment of a sentence and thereby becomes comprehensible to us.”  (Frege is using “thought” here to refer to a proposition, i.e. to the content of a “thought” in the mentalistic sense of the term.)

Now unlike Frege, Aristotle and Aquinas are not Platonic realists.  But they are moderate realists, and they would affirm something like Frege’s basic point.  Not only the propositions we grasp in having thoughts, but the thoughts themselves, are immaterial and distinct from any visual or auditory images we might form of particular sentences.  But we nevertheless find it impossible to entertain a proposition, and thus to have a thought, without also forming either images of sentences or some other imagery.  And in the view of Aristotle and Aquinas, all imagery is, as I have said, bodily and thus material.  As Aquinas concludes in Book I, chapter 2 of his Commentary on Aristotle’s De Anima, “since one cannot have imagery without a material organ, it seems clear that there can be no intellectual operation without the cooperation of matter” (as translated by Robert Brennan at p. 192 of his Thomistic Psychology).

Hence the A-T theorist affirms that there will always be some material correlate to normal human intellectual activity – not as a reluctant concession forced on the theory by the successes of modern neuroscience, but, on the contrary, precisely as a prediction of the A-T position as it has been understood from the beginning.  Were Aristotle and Aquinas to be made familiar with the sorts of neuroscientific discoveries frantically trumpeted by materialists as if they should be an embarrassment to the dualist, they would respond, with a shrug: “Of course.  Told you so.”

What A-T denies, again, is that the neurological level of description, however necessary, can ever suffice to account for intellectual activity.  There will always in principle be some slack between the neuroscientific facts and the facts about the content of our thoughts – something even materialists like W. V. Quine and Donald Davidson have affirmed on philosophical grounds, and psychologists like Kagan have affirmed on empirical grounds.  For A-T, the main reason, as I have said, has to do with the contrast between the determinate and universal character of conceptual thought and the particular and indeterminate nature of material processes – see Ross’s article, linked to above, for an especially powerful presentation of this point.

This, incidentally, is why the A-T theorist is untroubled by the neuroscientific evidence for the possibility in principle of “mindreading,” which sometimes gets attention in the popular press.  Invariably, we are told that at least certain kinds of mental states can be “read off” the neurological evidence with a degree of accuracy that is both surprisingly high and yet considerably less than absolute.  For A-T, this is exactly what we should expect.  If a “phantasm” or image is material, so that we can in principle determine neurologically that you are entertaining such-and-such phantasms, then the circumstances under which you are doing so might make it likely that you are also entertaining thoughts of the sort typically associated with such phantasms.  But likelihood is the most we can ever attain given the slack between phantasms or imagery on the one hand, and conceptual content on the other – especially when the conceptual content abstracts considerably from anything we can imagine, as it does when we are thinking about matters far removed from what we can directly experience.

The fact is that Aristotelian-Thomistic hylemorphic dualism is the theory most clearly consistent with all of the philosophical and neuroscientific evidence.  Cartesian dualism is not refuted by such evidence, but it has to resort to arguably ad hoc measures in order to avoid certain difficulties (the interaction problem, the fact that we are sometimes completely unconscious, and so forth).  And there is absolutely nothing in the neuroscientific evidence to support reductive versions of materialism over against either property dualism or A-T.  In arguments for preferring materialistic reductionism to these dualist alternatives, all the work is being done by metaphysical and methodological assumptions rather than by empirical evidence – by bogus appeals to Ockham’s razor, say, or to the illusion that “everything else has been explained in materialist terms.”  (I say that the appeal to Ockham’s razor is in this context bogus, because the main arguments for dualism are not probabilistic “explanatory hypotheses” to which considerations of parsimony are relevant; they are, instead, attempts at strict metaphysical demonstration.  See the posts on Churchland linked to above for more on this issue.  And I say that the claim that “everything else has been explained in materialist terms” is an illusion for reasons set out here, here, and in the posts on Jackson and Chomsky linked to above.) 

Of course, property dualists, like A-T theorists, perceive that the mental and neurological levels of description are much closer than Cartesian dualists suppose; while non-reductive materialists like Davidson at least perceive that they are not as close as reductive materialists suppose.  But each of these views still suffers from analogues of the problems facing the more extreme versions of dualism and materialism.  For example, they both face the problem of epiphenomenalism, which follows upon their common “mechanistic” insistence that all causation be understood on the model of efficient causation.  Hylemorphic dualism is the true mean between the extremes, a view that has the advantages of the others without their difficulties.

So why are its virtues not more widely recognized?  The usual reasons:  There is, first of all, the average contemporary academic philosopher’s unfamiliarity with what the ancients and medievals really thought.  Second, there is the dogmatic, ideological status that the early moderns’ “mechanistic” revolution – their denial of Aristotelian formal and final causes – has taken on in modern intellectual life, bolstered by the wholly unmerited prestige that revolution has inherited from the successes of empirical science.  (See The Last Superstition for the details.)  And third, there is the equally dogmatic, equally ideological naturalism that sustains itself on the backs of the first two factors.  As Burge has written in another context:

The flood of projects over the last two decades that attempt to fit mental causation or mental ontology into a ‘naturalistic picture of the world’ strike me as having more in common with political or religious ideology than with a philosophy that maintains perspective on the difference between what is known and what is speculated.  Materialism is not established, or even clearly supported, by science.  (“Mind-Body Causation and Explanatory Practice,” in John Heil and Alfred Mele, eds., Mental Causation, at p. 117)